- Algorithm Creation - Creating any algorithm takes time. Creating error free algorithms takes a long time. The best
way to understand an algorithm is to grind through the process of errors and exceptions. If the structure is not right
try different solutions until the correct solution is found. Always logically work out the steps and there is always more
than one way to do things.
Starting from raw data is always the most challenging part of creating an algorithm. The data may or may not contain
any direct clues on how to systematically break it down into logic. The first step is to analyze the data and try
to decipher any patterns or commonalities. Many sets of data have common patterns that can be exploited.
Trying to create an algorithm while trying to incorporate each special case will result in an infinity of debugging.
To create an algorithm start with the largest case. This may be the most commonly used case or the case that
the most complex, however if it is to complex it should possibly be split up into smaller algorithms. If the purpose
of the algorithm cannot be summed in one sentence it should generally be divided apart.
When the main section of the algorithm is created test it until it functions properly. The simple steps are very easy to
debug and identify logic errors. Ideally trying to break the algorithm into smaller methods will create an algorithm
that is easier to follow and easier to run in a debugger.
- Current Projects -